Recursive feature elimination in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the Recursive Feature Elimination class from scikit-learn.

ML Python

from sklearn.feature_selection import [1]

Drag options to blanks, or click blank then click option'

ARFE

BPCA

CSelectKBest

DVarianceThreshold

Attempts:

3 left

💡 Hint

Common Mistakes

Confusing RFE with PCA which is for dimensionality reduction.

Using SelectKBest which is a different feature selection method.

The Recursive Feature Elimination class is called RFE in scikit-learn.

2fill in blank

medium

Complete the code to create an RFE object using a logistic regression model as the estimator.

ML Python

from sklearn.linear_model import LogisticRegression
rfe = RFE(estimator=[1], n_features_to_select=3)

Drag options to blanks, or click blank then click option'

ALogisticRegression()

BDecisionTreeClassifier()

CKNeighborsClassifier()

DRandomForestClassifier()

Attempts:

3 left

💡 Hint

Common Mistakes

Using a classifier other than logistic regression when the instruction specifies logistic regression.

Passing the class instead of an instance (missing parentheses).

RFE requires an estimator to rank features. Here, LogisticRegression() is used as the estimator.

3fill in blank

hard

Fix the error in the code to fit the RFE selector to data X and target y.

ML Python

rfe = RFE(estimator=LogisticRegression(), n_features_to_select=2)
rfe.[1](X, y)

Drag options to blanks, or click blank then click option'

Apredict

Btransform

Cfit

Dscore

Attempts:

3 left

💡 Hint

Common Mistakes

Using transform before fitting the selector.

Using predict or score which are for models, not selectors.

To train the RFE selector on data, use the fit method.

4fill in blank

hard

Fill both blanks to select features and transform the dataset using the fitted RFE selector.

ML Python

X_selected = rfe.[1](X)
selected_features = rfe.[2]_

Drag options to blanks, or click blank then click option'

Atransform

Bsupport

Cfit

Dpredict

Attempts:

3 left

💡 Hint

Common Mistakes

Using fit instead of transform to reduce features.

Using predict instead of support_ to get selected features.

Use transform to reduce X to selected features, and support_ to get a mask of selected features.

5fill in blank

hard

Fill all three blanks to create an RFE selector, fit it, and print the ranking of features.

ML Python

selector = RFE(estimator=[1], n_features_to_select=4)
selector.[2](X, y)
print(selector.[3]_)

Drag options to blanks, or click blank then click option'

ALogisticRegression()

Bfit

Cranking

Dtransform

Attempts:

3 left

💡 Hint

Common Mistakes

Using transform instead of fit to train the selector.

Trying to print support_ instead of ranking_ for feature ranks.

Create the selector with LogisticRegression(), fit it with fit(), and get feature rankings with ranking_ attribute.